Image de couverture de livre

Language models recognize dropout and Gaussian noise applied to their activations

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Fornasiere, Damiano, Bronzi, Mirko, Kitts, Spencer, Palmas, Alessandro, Bengio, Yoshua, Richardson, Oliver
Format:	Preprint
Publié:	2026
Sujets:	Artificial Intelligence
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Machine learning and information theory concepts towards an AI Mathematician
par: Bengio, Yoshua, et autres
Publié: (2024)

Can a Bayesian Oracle Prevent Harm from an Agent?
par: Bengio, Yoshua, et autres
Publié: (2024)

Superintelligent Agents Pose Catastrophic Risks: Can Scientist AI Offer a Safer Path?
par: Bengio, Yoshua, et autres
Publié: (2025)

Fast Monte Carlo Tree Diffusion: 100x Speedup via Parallel Sparse Planning
par: Yoon, Jaesik, et autres
Publié: (2025)

Efficient Causal Graph Discovery Using Large Language Models
par: Jiralerspong, Thomas, et autres
Publié: (2024)

Local Inconsistency Resolution: The Interplay between Attention and Control in Probabilistic Models
par: Richardson, Oliver E., et autres
Publié: (2026)

In-Context Parametric Inference: Point or Distribution Estimators?
par: Mittal, Sarthak, et autres
Publié: (2025)

GFlowNet Foundations
par: Bengio, Yoshua, et autres
Publié: (2021)

Rejecting Hallucinated State Targets during Planning
par: Zhao, Mingde, et autres
Publié: (2024)

Latent Veracity Inference for Identifying Errors in Stepwise Reasoning
par: Kim, Minsu, et autres
Publié: (2025)

A Complexity-Based Theory of Compositionality
par: Elmoznino, Eric, et autres
Publié: (2024)

In-Context Reinforcement Learning through Bayesian Fusion of Context and Value Prior
par: Berkes, Anaïs, et autres
Publié: (2026)

Monte Carlo Tree Diffusion for System 2 Planning
par: Yoon, Jaesik, et autres
Publié: (2025)

Shaping Inductive Bias in Diffusion Models through Frequency-Based Noise Control
par: Jiralerspong, Thomas, et autres
Publié: (2025)

Geometric Signatures of Compositionality Across a Language Model's Lifetime
par: Lee, Jin Hwa, et autres
Publié: (2024)

Generative Recursive Reasoning
par: Baek, Junyeob, et autres
Publié: (2026)

Structure-Aligned Protein Language Model
par: Chen, Can, et autres
Publié: (2025)

Were RNNs All We Needed?
par: Feng, Leo, et autres
Publié: (2024)

FALCON: Few-step Accurate Likelihoods for Continuous Flows
par: Rehman, Danyal, et autres
Publié: (2025)

Distributional GFlowNets with Quantile Flows
par: Zhang, Dinghuai, et autres
Publié: (2023)

Expert-Guided LLM Reasoning for Battery Discovery: From AI-Driven Hypothesis to Synthesis and Characterization
par: Liu, Shengchao, et autres
Publié: (2025)

Active Attacks: Red-teaming LLMs via Adaptive Environments
par: Yun, Taeyoung, et autres
Publié: (2025)

Chapter Monitoring online perception of environmental issues on coasts of Sicily
par: De Marchi, Damiano, et autres
Publié: (2022)

Divergent Creativity in Humans and Large Language Models
par: Bellemare-Pepin, Antoine, et autres
Publié: (2024)

Learning What Matters: Steering Diffusion via Spectrally Anisotropic Forward Noise
par: Scimeca, Luca, et autres
Publié: (2025)

Discrete, compositional, and symbolic representations through attractor dynamics
par: Nam, Andrew, et autres
Publié: (2023)

Consciousness-Inspired Spatio-Temporal Abstractions for Better Generalization in Reinforcement Learning
par: Zhao, Mingde, et autres
Publié: (2023)

Action abstractions for amortized sampling
par: Boussif, Oussama, et autres
Publié: (2024)

Diffusion Generative Flow Samplers: Improving learning signals through partial trajectory optimization
par: Zhang, Dinghuai, et autres
Publié: (2023)

Trees and spectra of Heyting algebras
par: Fornasiere, Damiano, et autres
Publié: (2024)

Intuitionistic Sahlqvist theory for deductive systems
par: Fornasiere, Damiano, et autres
Publié: (2022)

Can Safety Fine-Tuning Be More Principled? Lessons Learned from Cybersecurity
par: Williams-King, David, et autres
Publié: (2025)

Mitigating Shortcut Learning with Diffusion Counterfactuals and Diverse Ensembles
par: Scimeca, Luca, et autres
Publié: (2023)

Trajectory Flow Matching with Applications to Clinical Time Series Modeling
par: Zhang, Xi, et autres
Publié: (2024)

Assessing SAM for Tree Crown Instance Segmentation from Drone Imagery
par: Teng, Mélisande, et autres
Publié: (2025)

Adaptive Inference-Time Scaling via Cyclic Diffusion Search
par: Lee, Gyubin, et autres
Publié: (2025)

Unlearning via Sparse Representations
par: Shah, Vedant, et autres
Publié: (2023)

Vision language models have difficulty recognizing virtual objects
par: Tran, Tyler, et autres
Publié: (2025)

AIMS.au: A Dataset for the Analysis of Modern Slavery Countermeasures in Corporate Statements
par: Bora, Adriana Eufrosina, et autres
Publié: (2025)

Efficient Regression-Based Training of Normalizing Flows for Boltzmann Generators
par: Rehman, Danyal, et autres
Publié: (2025)